🚀 go-pugleaf

When a CMISE m-get-indication is received, a performing CMISE user (the
agent) can either generate a single m-get-response or generate multiple
m-get-responses (according to X.710 | ISO 9595, section 8.3.1.2).

To my understanding, the "single" procedure is suitable when a single
managed object (MO) is selected as a result of scoping and filtering.
The "multiple" procedure is suitable when more than one MO is thus
selected.

The single procedure works as follows. The performing CMISE user
generates a single m-get-response without a linked identifier but with
an attribute list.

The multiple procedure works as follows. The performing CMISE user
generates an m-get-response for each selected MO with a linked
identifier and with an attribute list. After the last one, the
performing CMISE user sends an extra m-get-response, without a linked
identifier and without an attribute list.

My question is: when to use the single procedure and when the
multiple?

Of course the multiple procedure must be used if there is more than one
MO. But if there is exactly one MO, both procedures are suitable. Is
there a rule which defines what to do in this case?  One could imagine
a rule saying: "If there is exactly one MO selected by scoping and
filtering, the single procedure shall be used". Or there could be a
rule: "If scope=base object only, the single procedure shall be used,
and otherwise the multiple procedure". But I haven't found such a rule
yet.
-- 
 __    __     Charles van den Ouweland              Alcatel Telecom Systems
/   V /  \    ouweland%31.40.dnet@se.alcbel.be                P.O. Box 3292
\__ D \__/    Phone: +31 70 3079 126                       2280 GG Rijswijk
              Fax:   +31 70 3079 274                        the Netherlands